Subject:
|
Not that you care...
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 6 Aug 1999 19:21:34 GMT
|
Original-From:
|
John A. deVries II <zozzles@lanl.gov=spamcake=>
|
Viewed:
|
1087 times
|
| |
|
|
If anyone isn't unsure, I concede that I goofed with respect to the word
"fluid". I still stand staunchly by the distinction between "hydraulic"
and "pneumatic".
Apologies and all that to whom they are appropriate.
-----------------------------------
A small introduction:
I'm one of those middle-aged electrical/software engineer types who work at
the Los Alamos National Laboratory (much in the news lately). Near the
beginning of this decade there were a couple of articles in the most
mundane of the I.E.E.E. publications about the LEGO robotics projects that
were in progress at MIT. Having been hooked on robotics since I was about
8 years old, I started buying LEGO...
I now have somewhere between $2000 and $3000 worth of the stuff. I kept
finding that the Newest Kit(s) had parts that previous ones didn't and that
they certainly looked like I sure needed to have them. This wasn't
entirely a compulsive disorder, however: I've built the old Super Car and
the big pneumatic truck amongst a lot of the other kits. Whilst collecting
all of this stuff I never got around to building a HandyBoard or such stuff
-- in fact, I've been much more a part of the BEAM robotics schtick (see,
for example, http://www.serve.com/heretics).
Now that I have a MindStorms kit, I expect to finally make some progress in
the area of robotics & make a decent contribution to this list rather than
merely be a gadfly. In particular, I've got this suspicion that mapping
needs a new paradigm, one that -expects- the map to change (you've moved
the furniture, dropped a book, someone blew up the mountain <grin>, etc.)
If anyone can direct me to some fresh ideas, I'd be most interested.
Somewhat along those lines & realizing that the amount of memory in the RCX
is quite limited, is there some way to create "persistent objects" (well,
even storing a couple variables would be interesting) using the standard
firmware/Spirit? I don't see anything that looks like such a thing in the
RC command listing, particularly not one that is going to survive a given
program being stopped or the RCX being powered down (I don't expect the
timer value to be of much help here). If not, I presume that legOS or
hforth or Jini/MindStorms or any of the other firmware-replacements can do
such a thing: after all, the firmware itself is persistent.
BTW: the link to
http://hitachi.eps.bertelsmann.de/hitadb/micro/getdoclist.html?key=16
(H8/300 manuals) seems to be broken (at least just this moment -- just
obeying the suggestion to check the web site first, y'see) otherwise I'd
have a better idea how memory is organized in the RCX processor. It seems
that one of the things that would most greatly affect the usefulness of an
RCX-like product would be the amount of persistent/total memory there is
available. I realize that having more or different inputs/outputs and the
nature of the processor itself are big factors, but if those happen to be
fixed then memory has -got- to be the biggy as far as "intelligence" or
"ability" goes.
This could easily be a part of the dreaded upgrade & as a result wouldn't
affect much else...
Zoz
---------------------------------------------------------------
John A. deVries II
zozzles@lanl.gov
--
Did you check the web site first?: http://www.crynwr.com/lego-robotics
|
|
Message has 3 Replies:
 | | Re: Not that you care...
|
| (...) RCX (...) The only way I can think of is to fake it. Using the API approach, there's no reason why you couldn't define some persistent data that your code (in VB, C++ or whatever) would write to disk as appropriate. The idea is that a (...) (25 years ago, 6-Aug-99, to lugnet.robotics)
| | | Re: Not that you care...
|
| "John A. deVries II" wrote: ... (...) I hope you'll post anything intersting you find/invent to the newsgroup, since I am interested in that too and I believe there are more of us. when I get around to actually build some robot, I plan to make a big (...) (25 years ago, 9-Aug-99, to lugnet.robotics)
| | | Re: Not that you care...
|
| (...) This isn't exactly the type of adaptive mapping you're talking about, but it will give you an idea of the complexity of the task. Go to: (URL) and read the sections on "occupancy maps" and "localization". (...) Right. No persistent variables (...) (25 years ago, 10-Aug-99, to lugnet.robotics)
|
Message is in Reply To:
| | Darn those definitions (was: The new Super Car)
|
| (...) Wrong-o. Sorry to be flameish about this, but generally speaking (approximately) four states of matter are accepted: solid, liquid, gas and plasma. Liquids and gases have fairly distinct differences having to do with the degree of (...) (25 years ago, 6-Aug-99, to lugnet.robotics)
|
15 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
